Big Data on the High Seas

Andy McKeran and his team at GE build technology for drill ships, blowout preventers, risers and other machines battered by the seas as far as 100 miles from the coast for months at a time. When something goes wrong, it might take a while to fix it.

Outages to the 'money line' that's tapping subsea oil reservoirs, for example, can quickly add up to $700,000 per day. "A single failure in the chain of equipment can cause an unpredictable and unproductive event," McKeran explained. "What we need is a 'God's view' of the entire system and the ability to prevent problems before they strike."

That view is called SeaStream. It allows drilling companies to see the vessel and the subsea equipment as a whole, monitor it in real time, record and analyse its history and search for anomalies. "It could take you up to three weeks to find and fly the right expert to a remote platform to fix something," McKeran said. "But software and connectivity can actually bring the issue to the experts on shore."

Over the last few years, power plants, airlines, hospitals and other businesses have started using big data systems built on GE's Predix software platform. The company's software engineers developed it specifically for the Industrial Internet, a digital network connecting, collecting and analysing data from a myriad of sensors already installed in locomotives, jet engines and other intelligent machines.

GE built SeaStream to be 'manufacturer-agnostic', meaning it can work with machines of any make. McKeran says that SeaStream could help reduce unplanned downtime by 20% and third-party costs associated with repairs by a quarter, or in the region of $16million per vessel per year.